Persistent Organic Pollutants in Fish Oil Supplements on the Canadian Market: Polychlorinated Biphenyls and Organochlorine Insecticides

Authors

  • Dorothea F.K. Rawn,

    1. Author Rawn is with Food Research Div., Bureau of Chemical Safety and author Feeley is with Chemical Health Hazard Assessment Div., Health Products and Food Branch, Health Canada, Sir Frederick Banting Research Centre, 251 Sir Frederick Banting Driveway, Tunney's Pasture, Ottawa, ON, Canada, K1A 0K9. Authors Breakell, Verigin, Nicolidakis, and Sit are with Food Directorate, Western Region, 3155 Willington Green, Burnaby, BC, Canada, V5G 4P2. Direct inquiries to author Rawn (E-mail: Thea_Rawn@hc-sc.gc.ca).
    Search for more papers by this author
  • K. Breakell,

    1. Author Rawn is with Food Research Div., Bureau of Chemical Safety and author Feeley is with Chemical Health Hazard Assessment Div., Health Products and Food Branch, Health Canada, Sir Frederick Banting Research Centre, 251 Sir Frederick Banting Driveway, Tunney's Pasture, Ottawa, ON, Canada, K1A 0K9. Authors Breakell, Verigin, Nicolidakis, and Sit are with Food Directorate, Western Region, 3155 Willington Green, Burnaby, BC, Canada, V5G 4P2. Direct inquiries to author Rawn (E-mail: Thea_Rawn@hc-sc.gc.ca).
    Search for more papers by this author
  • V. Verigin,

    1. Author Rawn is with Food Research Div., Bureau of Chemical Safety and author Feeley is with Chemical Health Hazard Assessment Div., Health Products and Food Branch, Health Canada, Sir Frederick Banting Research Centre, 251 Sir Frederick Banting Driveway, Tunney's Pasture, Ottawa, ON, Canada, K1A 0K9. Authors Breakell, Verigin, Nicolidakis, and Sit are with Food Directorate, Western Region, 3155 Willington Green, Burnaby, BC, Canada, V5G 4P2. Direct inquiries to author Rawn (E-mail: Thea_Rawn@hc-sc.gc.ca).
    Search for more papers by this author
  • H. Nicolidakis,

    1. Author Rawn is with Food Research Div., Bureau of Chemical Safety and author Feeley is with Chemical Health Hazard Assessment Div., Health Products and Food Branch, Health Canada, Sir Frederick Banting Research Centre, 251 Sir Frederick Banting Driveway, Tunney's Pasture, Ottawa, ON, Canada, K1A 0K9. Authors Breakell, Verigin, Nicolidakis, and Sit are with Food Directorate, Western Region, 3155 Willington Green, Burnaby, BC, Canada, V5G 4P2. Direct inquiries to author Rawn (E-mail: Thea_Rawn@hc-sc.gc.ca).
    Search for more papers by this author
  • D. Sit,

    1. Author Rawn is with Food Research Div., Bureau of Chemical Safety and author Feeley is with Chemical Health Hazard Assessment Div., Health Products and Food Branch, Health Canada, Sir Frederick Banting Research Centre, 251 Sir Frederick Banting Driveway, Tunney's Pasture, Ottawa, ON, Canada, K1A 0K9. Authors Breakell, Verigin, Nicolidakis, and Sit are with Food Directorate, Western Region, 3155 Willington Green, Burnaby, BC, Canada, V5G 4P2. Direct inquiries to author Rawn (E-mail: Thea_Rawn@hc-sc.gc.ca).
    Search for more papers by this author
  • M. Feeley

    1. Author Rawn is with Food Research Div., Bureau of Chemical Safety and author Feeley is with Chemical Health Hazard Assessment Div., Health Products and Food Branch, Health Canada, Sir Frederick Banting Research Centre, 251 Sir Frederick Banting Driveway, Tunney's Pasture, Ottawa, ON, Canada, K1A 0K9. Authors Breakell, Verigin, Nicolidakis, and Sit are with Food Directorate, Western Region, 3155 Willington Green, Burnaby, BC, Canada, V5G 4P2. Direct inquiries to author Rawn (E-mail: Thea_Rawn@hc-sc.gc.ca).
    Search for more papers by this author

Abstract

ABSTRACT:  Fish and seal oil dietary supplements, marketed to be rich in omega-3 fatty acids, are frequently consumed by Canadians. Samples of these supplements (n = 30) were collected in Vancouver, Canada, between 2005 and 2007. All oil supplements were analyzed for polychlorinated biphenyls (PCBs) and organochlorine insecticides (OCs) and each sample was found to contain detectable residues. The highest ΣPCB and ΣDDT (1,1,1-trichloro-di-(4-chlorophenyl)ethane) concentrations (10400 ng/g and 3310 ng/g, respectively) were found in a shark oil sample while lowest levels were found in supplements prepared using mixed fish oils (anchovy, mackerel, and sardine) (0.711 ng ΣPCB/g and 0.189 ng ΣDDT/g). Mean ΣPCB concentrations in oil supplements were 34.5, 24.2, 25.1, 95.3, 12.0, 5260, 321, and 519 ng/g in unidentified fish, mixed fish containing no salmon, mixed fish with salmon, salmon, vegetable with mixed fish, shark, menhaden (n = 1), and seal (n = 1), respectively. Maximum concentrations of the other OCs were generally observed in the seal oil. The hexachlorinated PCB congeners were the dominant contributors to ΣPCB levels, while ΣDDT was the greatest contributor to organochlorine levels. Intake estimates were made using maximum dosages on manufacturers' labels and results varied widely due to the large difference in residue concentrations obtained. Average ΣPCB and ΣDDT intakes were calculated to be 736 ± 2840 ng/d and 304 ± 948 ng/d, respectively.

Introduction

Fish consumption is associated with health benefits in part due to the presence of the essential polyunsaturated fatty acids (PUFAs), particularly the omega-3 fatty acids in fish, which cannot be synthesized by mammals (Hamilton and others 2005). The other essential lipids, omega-6 fatty acids, are found in fish tissue; however, they are also available through other foods (Hamilton and others 2005). The intake of the omega-3 fatty acids in fish is associated with improvement in vascular health and reduced risk of coronary heart disease in humans (Storelli and others 2004; Covaci and others 2007). Additionally, consumption of fish has been linked to both a reduction in incidence of diabetes and symptoms of rheumatoid arthritis (Storelli and others 2004). As a result of the health benefits, Canadians continue to be encouraged to eat fish on a frequent basis (for example, 2 fish servings weekly) (Health Canada 2007a).

Over the last decade, elevated levels of persistent organic pollutants (POPs), including polychlorinated biphenyls (PCBs) and organochlorine insecticide (OC) compounds, have been reported in fish products around the world, particularly in farmed fish rather than those collected from the wild (Easton and others 2002; Hites and others 2004). In contrast to the PUFAs present in fish oils, POPs have been associated with negative health implications including immunotoxicity and carcinogenicity (Fernandez and others 2004). Impacts to normal endocrine function and reproduction also have been observed in fish and other wildlife, leading to concerns regarding potential impacts to humans (Safe 2000; Safe 2004; Sapozhnikova and others 2004). Public concern over potential negative health impacts due to contaminant exposure has resulted in follow-up surveys and assessments to establish the level of risk associated with consumption of retail fish/seafood in Canada. The results of this study indicated that the levels of POPs in fish did not pose a human health risk (Tittlemier and others 2004; Rawn and others 2006; Health Canada 2007b).

Because the public remains interested in enjoying the health benefits associated with the consumption of food rich in essential fatty acids, individuals consume fish oil supplements to enhance their health. PCBs and OC insecticides are lipophilic compounds and, therefore, are anticipated to be present in fish oils. The preparation of fish oil supplements, however, does include the processing of raw fish oils to remove specific contaminants including trace elements, pigments, odors, oxidation products, and so on, and can also result in a corresponding reduction in the POP levels (Hilbert and others 1998). Reductions can range from 0% to 100% depending on the refinement step (Hilbert and others 1998). Comparative testing of different grades of fish oil, however, has shown that POP concentrations in fish oils vary widely and do not always correspond to the marketing grade (Jacobs and others 1998).

POP concentrations in fish oils vary due to the type of fish used in the oil preparations as well as the location from which the fish were collected (Falandysz and others 1994; Jacobs and others 2004; Akutsu and others 2006). For example, cod liver oils have elevated POP levels relative to those measured in other fish oils or fish oils mixed with vegetable oils (Jacobs and others 2004). Although POP levels in biota are readily available in the literature, concentrations in fish-based oils are not reported as frequently.

A wide variety of fish oils are available on the Canadian retail market and can be consumed as part of an individual's daily health regimen. Individual supplements may contain oil from a single fish, marine mammal (for example, seal), or a mixture of oils from more than 1 type of fish. Additionally, some supplements contain vegetable or grain oils combined with fish oil. This study was developed to evaluate the concentrations of POPs in dietary supplements containing fish-based oils marketed as sources of omega-3 fatty acids, available in on the Canadian market. Samples were collected between 2005 and 2007, with the analyses completed in 2007.

Materials and Methods

Samples

Fish oil supplements were collected from pharmacies, health food stores, Internet distributors, and large supermarkets in Vancouver, Canada. The supplements selected included those that were available locally, via small suppliers, in addition to those distributed across Canada to ensure collection of a wide range of categories of these products. A total of 30 samples (Table 1) were collected for analysis and included supplements containing oil from individual species (for example, salmon, seal, or shark) or mixtures of oil from a variety of fish sources (for example, salmon, anchovy, sardine, mackerel). Although oil type was identified on sample containers in most cases, the geographical location from which the fish, marine mammals, or vegetables were collected was not available. Some supplements contained fish oil mixtures, but species types were not identified (n = 3) and 2 samples were mixed with unspecified vegetable oils.

Table 1—.  Range of PCB concentrations [median] (ng/g) in oil supplements by type.
TypenMaximum recommended dosage (g)ΣPCBsaΣMarker PCBsb
  1. aΣ of PCB 1, 3, 4/10, 6, 8, 19, 18, 15, 16, 54, 31, 28, 33, 22, 52, 49, 104, 44, 37, 41, 40, 74, 70, 66, 95, 155, 60, 84/90/101, 99, 119, 97, 87, 81, 85, 110, 77, 151, 135, 123, 149, 118, 114, 188, 153, 168, 105, 141, 137, 138, 158, 178, 129, 126, 187, 183, 167, 128, 174, 177, 202, 171, 156, 157, 201, 180, 193, 191, 200, 169, 170, 199, 203, 189, 208, 207, 194, 205, 206, 209.

  2. bΣMarker PCBs: 28, 52, 101, 118, 138, 153, 180.

No identification33 to 65.84 to 40.9 [34.5]2.57 to 18.9
Mixed (no salmon)81.2 to 3  0.711 to 37.9 [24.2] 0.227 to 14.4 
Mixed (including salmon)63 to 619.3 to 26.5 [25.1]8.88 to 12.3
Salmon72 to 636.1 to 170 [95.3] 12.7 to 54.8
Vegetable and mixed21.2 to 2.77.70 to 16.3 [12.0]3.30 to 7.63
Shark21.581.0 to 10400 [5260] 38.6 to 5080
Menhaden13321104
Seal13519295

Chemicals and materials

All solvents (acetone, cyclohexane, dichloromethane [DCM], hexane, nonane, and toluene) used in sample and standard preparation were distilled in glass grade (Omnisolve, EM Science, Gibbstown, N.J., U.S.A.). Prior to use, glass wool and glassware were rinsed with acetone and hexane. Florisil (60 to 100 mesh) (Fisher Scientific, Hampton, N.H., U.S.A.) was activated at 300 °C for 7 h and silica gel (100 to 200 mesh) was supplied by Sigma-Aldrich (Oakville, ON, Canada). Activated carbon (Carbopack C) was purchased from Supelco (Oakville, ON, Canada), while celite was supplied by Fisher Scientific. Glass wool, celite, silica, sodium sulfate, and Florisil were all Soxhlet extracted for several hours with DCM and subsequently heated overnight to 300 °C prior to use. They were stored in a desiccator until use. Trace pure grade anhydrous sodium sulfate was purchased from Baker (Cheshire, U.K.), and heated at 300 °C for 18 h prior to use.

Analytical standards of both the native and 13C analogues of the PCBs and OC insecticides were purchased from Wellington Laboratories (Guelph, ON, Canada). All standards were prepared as mixtures in toluene or nonane.

Extraction and clean up

The oil from individual capsules of a given product were removed from the capsules and combined. Subsamples of the oil (0.5 g) were spiked with surrogate standards containing 35 13C PCBs (1, 3, 4, 8, 15, 19, 28, 52, 54, 70, 77, 81, 95, 101, 104, 105, 114, 118, 123, 126, 128, 138, 153, 155, 156, 157, 167, 169, 170, 180, 189, 202, 205, 208, 209) and 13C analogues of all the OC compounds measured (1,2,3,4 tetrachlorobenzene, pentachlorobenzene, α-HCH, β-HCH, γ-HCH, hexachlorobenzene, oxychlordane, trans-chlordane, trans-nonachlor, cis-nonachlor, p,p′-DDE, p,p′-DDT, and mirex) with the exception of 1,2,3,5 tetrachlorobenzene, prior to extraction. The subsample was diluted to 3 mL in a 15-mL centrifuge tube with DCM.

The initial step in lipid removal was performed using an Agilent high performance preparative gel permeation chromatography (GPC) system (New Castle, Del., U.S.A.), consisting of an 1100 Series quaternary pump, autosampler, and a fraction collector. Two Envirogel columns (150 × 19 mm and 300 × 19 mm) (Waters, Milford, Mass., U.S.A.) were used in series and DCM as the mobile phase with a flow rate of 5 mL/min. The GPC eluate was evaporated to near dryness and was reconstituted in 1 mL hexane.

Following the GPC step, 10 mL disposable serological pipettes plugged with glass wool were filled with 4 g silica gel that had been treated with ACS grade sulphuric acid (75 g silica gel: 50 g H2SO4) (EM Science) and topped with anhydrous sodium sulfate. Prior to adding a sample to the column bed, 4 × 5 mL volumes of hexane were used to condition the column. After the samples were quantitatively transferred to the silica column, the sample flask was rinsed with 3 × 5 mL volumes of hexane and added to the top of the column. Individual columns were then eluted with 50 mL hexane and the eluate was taken down to approximately 1 mL using rotary evaporation.

Sample extracts were cleaned up further on 5 mL serological pipettes filled with activated Florisil (1.5 g) and topped with sodium sulfate. The columns were preconditioned with 2 × 5 mL DCM followed by 2 × 5 mL hexane. Samples were quantitatively transferred to the top of the column bed and washed onto the column with 4 × 1 mL hexane and the eluate discarded. The sample flasks were then rinsed 3 times with 4 mL DCM, which was then added to the top of the column bed, and the eluate was collected in a 15-mL centrifuge tube. This fraction, containing organochlorines and most PCBs, was concentrated using gentle blowing with ultra-high purity nitrogen using 200 μL nonane as a keeper. These extracts were reconstituted in approximately 1 mL hexane and cleaned up on a 2nd Florisil column as previously mentioned.

This fraction was then evaporated, as before, to 200 μL using nonane as a keeper and then transferred to 700 μL autosampler vials with 2 × 200 μL DCM and evaporated to near dryness, but the extracts were never allowed to go to complete dryness. Finally, the extracts were reconstituted in 25 μL of recovery standard containing 13C PCB 9, 37, 79, 111, 162, 194, and 206.

Analysis

Both PCBs and OC compounds were analyzed using a Waters Autospec Premier high-resolution mass spectrometer (Manchester, U.K.) linked to an Agilent 6890 gas chromatograph (GC) (Palo Alto, Calif., U.S.A.) equipped with a splitless injection system. The column used for the GC separation was a fused silica DB-5 60 m × 0.25 mm × 0.25 μm (J & W Scientific, Folsom, Calif., U.S.A.) with a 1 m × 0.53 mm (J & W Scientific) retention gap. The injector temperature was set to 300 °C for all analyses, with a purge time of 1.5 min.

The initial oven temperature for the PCB analyses was 80 °C and held for 1.5 min, raised to 200 °C at 30 °C/min, followed by an increase to 280 °C at 5 °C/min, and held for 3 min. The oven temperature for the OC compounds was initially set to 80 °C and held for 1.5 min, raised at a rate of 30 °C/min to 200 °C, followed immediately by a 2nd increase in temperature to 280 °C at a rate of 5 °C/min, which was maintained for 3 min before the final temperature increase of 15 °C/min to 300 °C where it remained for 9 min, which completed the analytical run.

The electron energy was set to 37 eV, with a photomultiplier voltage of 350 V. The trap current was 650 μA and both the source and capillary line temperatures were maintained at 260 °C. The re-entrant temperature was 270 °C and perfluorokerosene-H (PFK) was used as the reference substance for tuning at 393 m/z. The mass resolution was set to 10000 for all analyses.

Quality assurance

The recovery of the 13C surrogate analogues was determined for each sample and the average recovery of surrogate standards ranged from 22% for PCB 209 to 92% for PCB 189. In addition to 13C PCB 209, 13C PCB 155 had average recoveries of <50% (42%). Average recoveries for the OC surrogates ranged from 30% to 102% for mirex and p,p′-DDT, respectively. The average recoveries of the 13C analogues of both 1,2,3,4 tetrachlorobenzene and pentachlorobenzene were low (40% and 45%, respectively). Analytes were corrected for surrogate recoveries.

A reagent blank and cod liver oil certified reference material (Natl. Inst. for Standards and Technology, Gaithersburg, Md., U.S.A.) was processed and analyzed with each set of samples prepared for PCB and OC analysis. Because PCB and OC concentrations in blank samples were very low relative to oil supplement samples, concentrations were not corrected for the level in blank samples. Analytes were present in the acceptable range based on certified concentrations.

The method detection limits (MDL) were established based on a 3: 1 signal to baseline noise ratio and are reported as averages of individual chromatograms, ranging from 1.39 to 8.61 pg/g and 2.52 to 9.28 pg/g for PCBs (104 and 40) and OC insecticides (HCB and p,p′-DDT) analyzed in this study, respectively.

The oil supplements tested in this study were found to have 100% lipid and, therefore, whole or wet weight was considered equivalent to lipid weight for these samples.

Mean, median, and standard deviation values were calculated using Microsoft Excel (2002; Redmond, Wash., U.S.A.). Correlations were determined using the Mann–Whitney Rank Sum Test with SigmaStat for Windows version 3.11 (2004; San Jose, Calif., U.S.A.).

Results and Discussion

Concentrations

PCBs and OC insecticides were detected in all supplement samples collected. Lowest ΣPCB concentrations (0.711 ng/g) were observed in a supplement that was comprised more than 1 fish oil but had no salmon or vegetable content. In contrast, the highest ΣPCB levels (10400 ng/g) were observed in a supplement identified as containing shark oil (Figure 1). Although this 1 shark sample had PCB concentrations exceeding 10000 ng/g, total PCBs in the 2nd shark oil supplement tested was 81 ng/g. The ΣPCB concentration observed in the 2nd sample more closely reflects the results reported by other researchers where PCBs in shark liver oils ranged in concentration from 16 to 340 ng/g (Akutsu and others 2006). The production differences in the 2 shark samples are expected to be minimal because they were marketed by the same company. The 2 samples, however, were from different lots and, therefore, are expected to have been prepared using different sharks. Differences in shark species, gender, geographical region, and season of collection contribute to a wide variability in POP levels.

Figure 1—.

Total PCB concentrations in oil supplement samples grouped by oil type (veg = vegetable).

In a limited sample (n = 6) of fish oil supplements available on the Swiss market, total PCBs (reported as ΣPCB congeners 28, 52, 101, 138, 153, and 180) ranged from 0.23 to 17 ng/g (Zennegg and Schmid 2006). In comparison, the sum of marker PCB congeners (ΣPCB 28, 52, 101, 118, 138, 153, 180) from the present survey of fish oil containing supplements, exclusive of the menhaden, seal, and shark oils, ranged from 0.227 to 54.8 ng/g, while results from a survey of fish oil supplements collected in the United Kingdom (n = 33) reported marker PCB concentrations of only 0.008 to 0.267 ng/g (mean value 0.098 ng/g) (Fernandes and others 2006). Shim and others (2003) reported mean total PCB concentrations in 26 oil supplement samples, 24 containing fish oils and 2 comprising algae oils, with concentrations ranging from below detection limits to 276 ng/g. Although the present study resulted in detectable levels of PCBs in all samples, no algae or pure vegetable-based supplement samples were tested, limiting the comparison. PCB concentrations in the fish oil supplements in the present study, however, were found to be similar to those reported by Shim and others (2003) (Table 1).

Maximum OC concentrations frequently were observed in the single seal oil supplement analyzed relative to most other oil types tested. The shark oil sample that had very high PCB concentrations, however, also had the highest concentration of ΣDDT (sum of p,p′ DDT and p,p′-DDE) (3310 ng/g) and mirex (10 ng/g) of all supplements. Although residue data in seal tissue are available in the scientific literature (Schantz and others 1993; Tanabe and others 1994; Vorkamp and others 2008), data in seal oil supplements are scarce.

Median PCB concentrations, determined by oil type ranged from 12 ng/g in vegetable and mixed fish oils to 5260 ng/g in shark oils. With the exception of the menhaden, seal, and shark oil supplements, all median ΣPCB concentrations were below 100 ng/g (Table 1). Owing to the limited number of samples from these supplement categories, the data observed in this study may not accurately reflect the concentrations in oils of this type in general.

Levels of PCBs in salmon oil supplements ranged from 36.1 to 170 ng/g (Table 1) where 3 or approximately half of the samples tested had concentrations exceeding 100 ng/g. PCB concentrations observed in supplements prepared using salmon oil were consistent with the levels reported in salmon tissue samples from the Canadian market in 2002 (Rawn and others 2006). PCB concentrations in the salmon oils from the present study are lower than reported in a pharmaceutical grade salmon oil collected from the United Kingdom (Jacobs and others 1998). Marker PCBs were detected in salmon oils in the present study and ranged from 12.7 to 54.8 ng/g, which contrasts with the salmon oil originating from Norway/the United States reported by Jacobs and others (2004), where PCBs were below detection limits.

Reported marker PCBs (92 ng/g) in a menhaden fish oil supplement tested (Santillo 2007) were similar to the concentration observed in the present study (104 ng/g). The menhaden oil supplement tested in the present study had residues greater than those obtained in supplements containing anchovy, mackerel, sardine, and salmon oils, however, the maximum concentrations were observed in seal and shark supplements. Only the PCB concentrations in the menhaden, seal, and 1 shark oil sample would exceed the U.S. Council for Responsible Nutrition's recommended PCB standard for supplements containing fish oil (90 ng/g) based on marker PCB concentrations (Council for Responsible Nutrition 2006).

Similar to PCBs, OC levels in oil supplement samples varied widely between supplement types. Although individual contributors to ΣHCH (for example, α-HCH, γ-HCH) and Σchlordane (for example, oxychlordane) were below detection limits in a few samples, only 1 mixed fish oil sample, with no salmon content, was found to have nondetectable residues of ΣHCH (Table 2). Mirex also was below method detection limits in this sample, but was detected in all other samples tested in the present study. The shark oil sample found to have very high PCB residues was also found to have extremely high levels of DDE although the maximum concentration of most OC insecticides were observed in the seal oil analyzed (Table 2). The supplements containing seal, shark, and salmon oils had elevated levels of OC compounds relative to those containing other fish and vegetable oils (Table 2).

Table 2—.  Range of organochlorine concentrations [median] (ng/g) in oil supplements by type.
CategoryΣCBzaΣHCHbΣHCBΣChlordanecΣDDTdMirex
  1. aΣ of 1,2,3,4 tetrachlorobenzene, 1,2,3,5 tetrachlorobenzene, and pentachlorobenzene.

  2. bΣ of α-HCH, β-HCH, and γ-HCH.

  3. cΣ of oxychlordane, trans-chlordane, trans-nonachlor, and cis-nonachlor.

  4. dΣ of p,p′-DDT and p,p′-DDE.

No identification0.062 to 0.0810.011 to 0.5710.023 to 1.070.042 to 2.430.301 to 13.50.052 to 0.092
[0.067][0.065][0.088][0.252][4.23][0.066]
Mixed (no salmon)0.021 to 0.103ND to 0.8390.021 to 0.6280.011 to 0.6540.189 to 15.2ND to 0.044
[0.030][0.396][0.217][0.342][10.6][0.024]
Mixed (including salmon)0.037 to 0.7960.752 to 1.770.361 to 2.710.500 to 2.299.31 to 24.90.020 to 0.511
[0.047][0.836][0.597][0.636][11.1][0.042]
Salmon0.177 to 2.831.37 to 36.91.55 to 21.93.51 to 84.74.76 to 2500.099 to 3.53
[2.19][29.7][15.3][22.9][59.2][0.618]
Vegetable and mixed0.106 to 0.3070.448 to 1.570.174 to 1.220.224 to 0.9583.71 to 5.630.029 to 0.069
[0.206][1.01][0.697][0.591][4.67][0.049]
Shark0.058 to 0.1800.052 to 0.1012.52 to 3.405.92 to 121309 to 33103.40 to 10.0
[0.119][0.077][2.96][63.4][1810][6.72]
Menhaden0.0280.1530.0549.1428.90.309
Seal9.8479.052.12112671.27

Trends

In general, the hexachlorinated congeners contributed the greatest extent to total PCB levels, followed by the pentachlorinated and heptachlorinated congeners (Figure 2). Although the concentrations were high in shark oil, the congener profile in these samples was consistent with other oil types. Jacobs and others (2004) also found the hexachlorinated congeners to be the dominant contributors to total PCB levels in fish oil supplements.

Figure 2—.

Relative contribution of each homologue group to total PCBs by oil type (veg = vegetable).

In contrast to the observations of Jacobs and others (2002), β-HCH was the predominant HCH isomer (mean ratio β-HCH: ΣHCH = 0.643 ± 0.267) in the oil supplements in the present study. DDE was the predominant contributor to total DDT in both the present study and the study reported by Jacobs and others (2002). The DDE: ΣDDT ratio was determined to be 0.866 ± 0.145 in the present study.

A strong correlation between ΣPCB and ΣDDT concentrations in these supplement samples was found to occur (r2= 0.988), however, this was not the case for the other OC compounds. Jacobs and others (2004) similarly found a correlation between ΣPCB and ΣDDT concentrations in oil supplements. Although a weak correlation between total chlordane levels and ΣPCBs was observed (r2= 0.223), the correlations between ΣPCB concentrations and the other OC compounds analyzed were not found to be significant, with r2 values < 0.07.

Owing to the limited number of supplements containing similar types of oil in the present study, direct comparisons between the levels of PCBs and OC insecticides in 1 supplement type relative to another were not possible. PCB and OC concentrations in salmon oil (n = 6) supplements, however, were compared to the levels observed in all mixed fish oil supplements, where samples containing any mixtures of anchovy, mackerel, sardine, vegetable, and salmon oil were included in the category (n = 16) for these comparisons. PCB and OC insecticide levels were statistically significantly lower in mixed fish oils relative to salmon oils (PCB P < 0.001, HCB, mirex, chlorobenzenes, HCH, chlordanes, DDT; P < 0.001, P < 0.001, P < 0.001, P < 0.001, P < 0.001, P = 0.035) in the present study.

Intakes

Daily intake estimates were calculated by multiplying the concentration of contaminant in each supplement with the maximum recommended daily dosage following label directions. Total PCB intakes ranged between 0.896 ng/d (0.285 ng/d marker PCBs) and 15700 ng/d (7620 ng/d marker PCBs), in a supplement containing anchovy, mackerel, and sardines and a shark oil supplement, respectively. The seal oil supplement resulted in the next highest intake estimate (1560 ng/d). In comparison, consumption of a single 150 g meal of farmed salmon with an average ΣPCB concentration of 17.5 ng/g would result in exposure to 2630 ng PCBs. Of the OC insecticide related compounds, the highest mean intake was estimated for ΣDDT (304 ng/d) and mirex intake was calculated to be the lowest (1.58 ng/d) (Table 3). The intake estimates determined in the present study generally are quite similar to those reported in cod liver oil supplements (Storelli and others 2004). Total PCB intakes reported by Storelli and others (2004) ranged from 4 to 2000 ng/d in individual cod liver oils, while ΣDDT intakes ranged from 4 to 1240 ng/d. Mean intakes determined in the present study are within the values reported by Storelli and others (2004) (Table 3).

Table 3—.  Intake estimates following maximum suggested dosage and contaminant concentration (ng/d).
CompoundRangeAverage ± SD
ΣPCB0.896 to 15,600736 ± 2840
Marker PCB0.285 to 7620  345 ± 1380
ΣDDT0.265 to 4970  304 ± 948 
ΣHCHND to 237 22.9 ± 55.4 
ΣChlordane0.022 to 634   45.0 ± 123  
ΣCBz0.027 to 29.5  2.36 ± 5.74 
HCB0.030 to 156   14.6 ± 33.0 
MirexND to 15.01.58 ± 3.14 

Conclusions

Consumption of shark, seal, and menhaden oil supplements would contribute more to ΣPCB intake than other oil types, based on the results of this study. Seal oil consumption, however, would result in increased exposure to most OC insecticides analyzed, with the exception of ΣDDT and mirex. The mixed fish oils tested in the present study had lower PCB and OC insecticide levels than other oil types. Although a limited number of supplements were analyzed in this study, variable concentrations of PCBs and OCs were observed corresponding to oil type.

Ancillary